Sea Turtle Population Ecology
297
Because there are so many extrinsic and intrinsic factors affecting population
growth rates, it is difficult to assess the population-level impacts of our management
efforts. Long time lags prevent us from attributing an observed increase or decline
in abundance to a particular cause. For example, the recent increase in Kemp’s
ridleys is due to a combination of effects: nest protection, reduced trawling effort
in Mexico, and TED regulations, and possibly other changes in vital rates (Turtle
Expert Working Group, 1998; 2000). Egg protection efforts began in the 1960s and
intensified in the 1980s, yet population recovery did not really begin until the late
1980s. Empirical evidence suggests that more hatchlings are currently surviving to
maturity, likely because of an increase in survival rates afforded by TEDs (Heppell
et al., 2002a).
11.5 POPULATION MODELS AS TOOLS FOR TESTING
HYPOTHESES ABOUT POPULATION DYNAMICS
One of the key barriers to integration across scales in population biology has been
our failure to put all the factors that we think might influence the status of a
population into a common framework. Instead, we study what we know (e.g.,
behavior, physiology, toxicology, and fisheries interactions) and hope our hardfought insights can be brought to bear on understanding the dynamics of these
populations. This is particularly critical for research on threatened and endangered
species such as sea turtles. If factor A reduces fecundity by 10%, is that important?
Or, if factor B reduces juvenile survival by 5%, is that important? Is factor A more
important than factor B? We must put these factors into a common currency to
address these questions at the population level. Namely, we must estimate their ageor stage-specific effects on population vital rates (e.g., fecundity, growth, and survival). Integration of this sort is long overdue, and it is critical to forecasting the
effects of environmental change and human activities on the future of sea turtles
and other protected species.
Population models provide a useful framework to integrate what we know and
to clearly identify what we do not know. Analyses of these models also allow us
to compare management alternatives on the basis of their mode of action and
impact to determine which are likely to contribute most to population recovery
(Heppell et al., 2000b). Although the early models of sea turtle population dynamics were simplistic, we believe their qualitative results are robust for turtles
(Heppell, 1998), and generally are robust for long-lived species (Heppell et al.,
1999; 2000a). Model analyses not only identify poorly known parameters, but
indicate those factors expected to have the greatest impact on population growth
(Heppell et al., 2000b). In this way, research priorities can be more efficiently
focused. Finally, these models are constrained by model assumptions and the
available data; qualitative prediction based on life history constraints is generally
reliable, but precise quantitative prediction awaits additional data collection (Heppell et al., 2002b). Still, in the context of ecological forecasting, qualitative predictions have provided useful guidance to managers responsible for the recovery
of these endangered species.
1123 book.book Page 297 Tuesday, November 12, 2002 7:43 AM
297
Because there are so many extrinsic and intrinsic factors affecting population
growth rates, it is difficult to assess the population-level impacts of our management
efforts. Long time lags prevent us from attributing an observed increase or decline
in abundance to a particular cause. For example, the recent increase in Kemp’s
ridleys is due to a combination of effects: nest protection, reduced trawling effort
in Mexico, and TED regulations, and possibly other changes in vital rates (Turtle
Expert Working Group, 1998; 2000). Egg protection efforts began in the 1960s and
intensified in the 1980s, yet population recovery did not really begin until the late
1980s. Empirical evidence suggests that more hatchlings are currently surviving to
maturity, likely because of an increase in survival rates afforded by TEDs (Heppell
et al., 2002a).
11.5 POPULATION MODELS AS TOOLS FOR TESTING
HYPOTHESES ABOUT POPULATION DYNAMICS
One of the key barriers to integration across scales in population biology has been
our failure to put all the factors that we think might influence the status of a
population into a common framework. Instead, we study what we know (e.g.,
behavior, physiology, toxicology, and fisheries interactions) and hope our hardfought insights can be brought to bear on understanding the dynamics of these
populations. This is particularly critical for research on threatened and endangered
species such as sea turtles. If factor A reduces fecundity by 10%, is that important?
Or, if factor B reduces juvenile survival by 5%, is that important? Is factor A more
important than factor B? We must put these factors into a common currency to
address these questions at the population level. Namely, we must estimate their ageor stage-specific effects on population vital rates (e.g., fecundity, growth, and survival). Integration of this sort is long overdue, and it is critical to forecasting the
effects of environmental change and human activities on the future of sea turtles
and other protected species.
Population models provide a useful framework to integrate what we know and
to clearly identify what we do not know. Analyses of these models also allow us
to compare management alternatives on the basis of their mode of action and
impact to determine which are likely to contribute most to population recovery
(Heppell et al., 2000b). Although the early models of sea turtle population dynamics were simplistic, we believe their qualitative results are robust for turtles
(Heppell, 1998), and generally are robust for long-lived species (Heppell et al.,
1999; 2000a). Model analyses not only identify poorly known parameters, but
indicate those factors expected to have the greatest impact on population growth
(Heppell et al., 2000b). In this way, research priorities can be more efficiently
focused. Finally, these models are constrained by model assumptions and the
available data; qualitative prediction based on life history constraints is generally
reliable, but precise quantitative prediction awaits additional data collection (Heppell et al., 2002b). Still, in the context of ecological forecasting, qualitative predictions have provided useful guidance to managers responsible for the recovery
of these endangered species.
1123 book.book Page 297 Tuesday, November 12, 2002 7:43 AM